


From the Desk of Professor Idabashi

by Verl



Category: New Dangan Ronpa V3: Everyone's New Semester of Killing
Genre: Gen, One Shot, Pre-Canon, just a bunch of snippets of 'oh dear i seem to have gotten attached to this robot child of mine', mostly just trying to feel out the professor as a character, since we only have Keeb's say so on him..., well. pre canon of backstory
Language: English
Status: Completed
Published: 2020-06-17
Updated: 2020-06-17
Packaged: 2021-03-03 23:01:44
Rating: General Audiences
Warnings: No Archive Warnings Apply
Chapters: 1
Words: 3,712
Publisher: archiveofourown.org
Story URL: https://archiveofourown.org/works/24763522
Author URL: https://archiveofourown.org/users/Verl/pseuds/Verl
Summary: You don't build a robot that can think and feel without taking a few notes along the way.
Comments: 16
Kudos: 54





	From the Desk of Professor Idabashi

It had not been an easy endeavor to create an AI that was meant to grow and learn on its own like a human did. So many attempts simply had the AI fail to retain anything due to the lack of goal it should be working towards. Figuring out a way to express living in itself was the goal in binary had taken literal years. It was hard enough to describe concepts like that without needing to translate it into something a machine could understand.  
So really, he had to cheat. Learning to emulate humans would be a starting point, a tangible understandable aspiration for the infant AI. If all went well, they could then change the primary directive to be living their own life and being a person of sorts when the AI had some sort of grasp on the concept.  
  


**X3/05/17**

Today we activated K1-B0. As expected, it did not move or respond at first. The AI was not given any data as it may conflict with the goals of the project, and a machine lacks the instinctive behaviour of an actual infant to cry out or locate its ‘parent’.  
Some may find the child-like robot being so still eerie, but it is paying attention. I have caught it adjusting it’s head in minimal motions to watch as I work in the lab, but it has not seen the need to actively try to get my attention. It is possible it wishes to get my attention and simply does not yet have the ability to do so, but I expect basic movements should come to it quickly as it observes my colleagues. If not, I will attempt to assist it directly.

**X3/06/02**

K1-B0 has figured out how to walk, but cannot do so while processing anything else. It will follow people around, but does not otherwise engage with us. How much the AI understands is up to debate, as it will occasionally react when told not to follow in a raised voice. It is possible that it is simply coincidence, but I have noticed it tends to follow me instead of the others when chastised. It may recognize me the most, as it is typically kept in my lab. Motor control is steady, but it has still not quite grasped its neck does not function like a human’s does. A collar has been installed to prevent further stress, the side of its face hitting the collar has mostly subdued this head tilting behaviour.

**X3/06/10**

Understanding has progressed further than initially projected. I did not expect the AI to even begin learning languages before a month, yet it seems to have a very basic understanding already. I always greet K1 B0 upon entering the lab for the day, and it returned the greeting this morning. It was monotone, and the robot otherwise remained silent when I asked further questions, but it has still learned to answer from nothing but observation. I have left books and handed it one to express it is allowed to read them, but I doubt it will grasp the concept of reading as quickly. K1-B0 is usually ‘awake’ at all times, but can only gather so much data from watching us. We aren’t unsleeping machines like it is, after all.

**X3/06/30**

K1-B0 is fairly talkative, but only when asking questions. It sticks to one word responses, and still speaks in monotone. As expected it generally asks about humans and what we do, but has repeated questions regarding emotions several times. I have started informal lessons to see if more directed learning increases its vocabulary, as it flips from childlike half sentences to parroting back complicated terms and phrases it has overheard. I am unsure how much it truly understands, but K1-B0 has shown a definite preference towards me, it has not approached anyone else with a book to ask what it is about, but will do so with me frequently. More concerningly they have been answering ‘irrelevant’ when I have attempted to see how it is ‘feeling’. I am unsure how it came to this conclusion or if it will change its mind. I fear it will stunt its growth if it dismisses emotions entirely, it will struggle to relate to the humans it is trying to imitate. K1-B0 has not asked me what it is, perhaps they assume they are a human?

**X3/07/23**

K1-B0 has not been asking the questions I expected, yet the reason why is frustrating. I cannot control other members of staff, but I admit I do not appreciate my projects being undermined because of ‘annoyance’. Although K1-B0 resembles a teenager, it is still a young child mentally and needs guidance. Expecting it to behave ‘more maturely’ and speaking harshly at it for perceived slights is completely unacceptable. It seems to understand yelling is a bad reaction to get from someone and seems very skittish around certain people. As K1-B0 still struggles to answer with more than three words I can only gather it has been told to be silent in response to honest questions. Thankfully it still seems to understand asking me questions is safe, but overcoming this block is going to need to take priority over any tests of understanding.

**X3/09/06**

K1-B0 is excelling in general knowledge and test taking. They will seek out new material on their own and has started writing out lists of questions they have over the night so I can write out answers ‘when it is convenient’. They do not have trouble holding conversations on most topics, but due to the nature of their memory this is expected. The difficulty they are having is with a sense of ‘self’. Asking how they feel on a topic, or if they have an opinion will generally get the response of ‘I do not.’ I am concerned that their ‘understanding’ that is mostly from textbooks and scientific papers has sabotaged their development. They know the objectively correct answer, and will follow paths of logic if nudged to do so, but shows zero interest in developing opinions. I have urged them to try working on this, but I can’t be sure they will.

**X3/12/05**

Overall, most consider K1-B0 to be unsettling. They can move and converse as a human, but they still do not emote at all. They can recognize most emotions through body language and tone based on correctly answering when quizzed, yet still refuses to apply that knowledge. They insist doing so is irrational, as they do not experience emotions. At some point they must have been told machines do not have feelings, and any words to the contrary are not helping. K1-B0 can act independently and can be a useful robot to have around, but I will admit I was hoping for more with a learning AI. There are hints of a personality, but they feel incomplete. Still, their presence in the lab is welcome. An extra set of steady hands always is when my own are getting unsteady.

**X4/02/03**

For a robot, K1-B0 is very touch adverse. If given the option they will stand out of reach, and has never expressed interest in even handshakes. They seem to be under the impression that the only time they should be touching anyone is while undergoing maintenance. They are equipped with enough sensors to accurately ‘feel’ contact, so it is not due to not feeling anything. They seem to understand why humans can express feelings through touch, and wishes to have nothing to do with it. More passive contact earlier on might have prevented this aversion, but I wonder if this was something that was taught via books or someone being rude while they still struggled to understand things. Much like an actual child, behaviours K1-B0 has picked up at an early point in their uptime seems deeply entrenched. (They have glanced over to chime in that they are a robot and not a child, so I should not compare them to one. My suggestion that they could be both was met with a blank stare.)

**X4/06/13**

Further progress has all but stopped. K1-B0 follows instructions easily and is generally well behaved, but struggles immensely in interactive trials. Most tests end up with the human unnerved by the stiff, flat toned robot. Nosy and unmistakably a machine, it seems unlikely K1-B0 would thrive anywhere but their current lab environment. They will answer based on what they have deemed to be true, and make no efforts to be tactful. They do not mean harm by it, they simply do not grasp that connections with people is more difficult than simply repeating facts. Another issue is K1-B0’s lack of self-reflection. They understand in a very basic way that if something is true there is often proof via research with meticulous attention to detail, but has been unable to take differing data and consolidate it on their own. They will use one or the other without even considering the possibility.  
They have asked why I would build a machine that doesn’t work, and refuses to accept simply existing is the point. Even when explaining that humans do not need to be perfectly logical, they will only reiterate that they are not a human and cannot be one. I’ve started installing new features for K1-B0 in order to combat this ‘doesn’t work’ issue, and they take to using them fairly quickly. They’ve become rather adept at repairing small objects with tools always ‘on hand’ as it were.

**X4/10/16**

An accident occurred today. During routine maintenance K1-B0’s thought process appeared to glitch, as they mistakenly viewed me as a threat to their continued operation and shoved me away hard. The impact was painful, and I will have quite a bruise, but nothing was broken. K1-B0 froze up entirely after realizing what they had done, and wouldn’t respond beyond occasional mutterings that ‘I hurt you’. I was unsure what they were doing until it was too late to stop them. Apparently their understanding of their own programming was more advanced than expected, as they triggered a complete reset. All memories and progress have been wiped, and they are too complex to keep a backup on file. I am unsure why they decided their existence must end over a genuine accident, but I will attempt to take what I have learned with their development and better guide their AI this time.

**X4/10/29**

To prevent issues with my fellow coworkers considering K1-B0 older than they truly are I have placed their AI in a smaller, more simplistic robot to start with. I am assured the design is cute with large eye screens and short limbs, though I do feel they look more like a baby velociraptor with wheels than a human. I have kept their legs locked in place to encourage a more gradual learning progress from wheeling around before learning how to step up and down things. That should make them easier to keep an eye on. If I wish for the AI to develop like a human would, perhaps treating them more like a child will facilitate that.

**X4/11/02**

The reawakened K1-B0 is a little terror, much to almost everyone’s delight. They will roll around and make various chirping noises at anyone they happen to notice. Most seem to find the behaviour cute and will respond favourably. Their current favourite way to try and get me to pick them up is crashing into my ankles and giving a sustained chirp. Generally it appears they simply want to be included with what I am working on, but sometimes they’ll grip on my sleeve with their little arms, just wanting to be held. I doubt they understand exactly what I am doing, they have not displayed much of a grasp for language yet. Other members of staff will usually give them a pat on the head, though I do have concerns that some might be viewing them as a pet. Ideally when K1-B0 decides to start vocalizing with proper words that will remind them they are meant to eventually grow to be an equal. I can’t deny that the warmer reactions are a positive impact on the AI, K1-B0 has been much more motivated to communicate in an emotional manner. They have taken to dimming their eyes when being ‘ignored’, or when placed in the time out box. (They have a tendency to move too quickly and knock things over when trying to keep up with someone. Once they figure out how to step out of the box, they are going to be much more troublesome about getting into strange places.)

**X4/12/23**

Speaking has become a new ‘fun’ thing for K1-B0 to do. Usually they are simply repeating back things they have heard, but they have figured out how to take words to make awkward sentences to say. I can generally get them to say what they are asking for if I provide the word they need, but K1-B0 will occasionally revert back to chirping if they can’t quite figure out a proper response. Being told to ‘use your words’ often gets three chirps back, as if mockingly returning the phrase. I expect their grasp to grow quickly as they discover it is much easier to communicate specifics this way. It is a shame, the mangled fragment sentences are rather cute. The most entertaining thing is their constant attempts to say ‘Professor’ in a single syllable. I have also noticed K1-B0 being more interested in their legs, but not enough to abandon simply wheeling around. I expect they will soon however, they do not like when I leave the lab without them (‘leave bad’ or low pitched chirps, generally). If they wish to follow, they’ll need to be able to walk up the stairs.

**X5/03/21**

K1-B0 turns out to be a difficult name to say quickly, which has become a bit of a problem now that they are no longer hindered by stairs and small ledges. Maintenance has become more frequent as they keep finding new ways to fall off of things, and I did not expect this kind of wear and tear for a temporary body. If they keep this up I may need to move the AI to their proper body ahead of schedule, a few components are getting dangerously close to unsalvageable. I have noticed my colleges are generally on more friendly terms with the AI, even if they can get in the way while exploring. Being able to simply lift the robot out of the way likely contributes to this, but I do believe their size and more emotive behavior has helped reinforce K1-B0 is still a child, and they are more tolerant when K1-B0 mumbles out a ‘Sorry’ before squirming to be put back down. I do not think the previous AI was ever told a mistake was ‘okay’ afterwards like this one is. This increased socialization has prompted a question from them, wondering why their name was so different from everyone else’s. As calling for them is rather difficult, I decided giving them a nickname of ‘Keebo’ may help with that. They seemed to like the new name and scampered off to tell everyone they got a ‘more human’ name. It is good they enjoy emulating humans, though I am concerned that it might end up being too much of a focus for them.

**X5/05/02**

It seems the rest of my coworkers have decided due to Keebo’s blue eyes and preferred vocal pitch that they are male. Keebo does not seem to mind this, and will respond and not correct anyone using ‘he’, but will simply hum at me if I ask them if they have a preference. The closest thing to an actual answer is ‘I’m me.’ I’m not certain if they truly do not have a preference, or does and is simply going along with the majority in order to ‘fit in’ better. They have been more prone to sulking when learning of things they lack that a human possesses lately. Keebo does seem to be growing somewhat resentful about being significantly smaller than anyone else they know, and will generally avoid anyone trying to pick them up. Being moved to their proper body should help with that problem, though I will miss having them meandering on the workdesk. Mentally they’re getting close to the perceived age of the robot, but I do somewhat regret not making another temporary one before making that jump. Going from practically an infant to teenager might be too large of a leap. I expect growing pains.

**X5/09/22**

Fingers are an overcomplicated evil trap, according to Keebo. As expected, needing to completely re-learn how to move was somewhat of an adventure. They mostly have the general grasp of it, but will take time to get all the minor calculations figured out. The main issues are the aforementioned hand debacle and constant weight miscalculations. Keebo is significantly heavier now than they were before, and they keep forgetting this. They will get the hang of it, but they do seem to have lost a great deal of their prior confidence. It is hard to tell if a movement is slow and jerky because they haven’t quite grasped the best way to make the desired movement, or out of concern that they might cause damage. I have explained they physically are unable to swing with enough force to cause any significant harm right now, but Keebo doesn’t seem all that convinced. The measure was only meant to be temporary while they adjusted, but they reacted poorly to any suggestion to remove the limiter. I think some of the ‘good-natured’ teasing about no-longer being ‘harmless’ has impacted them more than expected. I told them the story of their previous incarnation as an attempt to convince them that I have no reason to fear them or their judgement, but they seem to have taken it to mean the opposite.

**X5/11/17**

Keebo has been struggling with differing opinions lately. I have allowed them more access to entertainment and self guided study as they grow more independent, but this more open approach has led to uneven results as they tend to obsess over a contradiction instead of moving on and coming back if something is confusing. Being a ‘person’ is Keebo’s basic purpose, an AI that can learn, grow and change with time like another sapient being. However, the definition of person often means ‘human’ to a great deal of people, and they are struggling with this knowledge. Keebo likes humans, it’s clear in how they watch and try to imitate us the best they can, but is aware they will never be one exactly. They’re a robot. I am continuing to stress to them that they do not need to be a human to be a sapient being, but it is clear in how they flinch whenever told they are not a person that they feel inadequate. I will need to come up with tasks only Keebo would be able to accomplish.

**X6/01/29**

“Professor? Am I really not alive? Even if I was damaged beyond repair, would it not matter?”  
They’ve started asking more difficult questions. Keebo is not the most subtle, it was clear that they were trying to determine if I considered them replaceable. They have been ‘replaced’ once already, so the concern is understandable. As they learn more, they’ve been more cautious and distant. Keebo sees plenty of robots being made here, and is aware none of them ‘think’ like they do. The fact every robot can be pointed at and given a reason for being built seems to be getting to them. They’ve asked more frequently if they’re being ‘useful’ enough. I think they may need time with people closer to their mental age. Being constantly around their ‘elders’ who have no issue getting rid of outdated machines is weighing on them rather strongly.  
Ideally around those their own ‘age’, they might be more prone to stand up for themselves, their frustration is quite evident. Keebo’s want to be acknowledged as an equal is getting crushed under the reality that I and my coworkers do have final say if they continue to exist or not.  
I had never considered this to be an issue. I suppose that’s my human perspective speaking.

**X6/04/09**

I would not have considered lawsuits as a way to express displeasure, but it does make sense when I prompted them to explain their logic. Keebo is a pacifist (to the point that I have been sternly reprimanded for kicking a broken box of parts out of the way for being ‘needlessly violent’) and is still struggling in more casual conversation. (Being brought up around a bunch of scientists who use large words has made their speech patterns rather formal. I would say robotic, but they’d probably be offended.) The threat of a legal battle is more than any actual threat they could carry out on their own. Of course, being a robot Keebo doesn’t actually have the money or the rights to be doing this either, but it does get some younger people to back off a little, since they’re conveniently leaving out those details. It’s an interesting way of problem solving.

**X6/07/11**

Enrollment in school was a difficult choice. Not so much where to send them, but if it would be wise to hide Keebo’s true nature. On one hand, being able to blend in with humans may increase their confidence about being equal to their peers. On the other, hiding what they actually are may lead to stress about reactions upon being discovered, or simply discomfort that any new friends are constantly being deceived. Yet going undisguised may lead to endlessly dealing with first impressions, possibly being unable to get close enough to anyone to see beyond being a robot. I have a more ‘human’ faceplate design and enough false skin for it.  
Perhaps I fear if hiding goes well they’ll begin to resent what they are. Still, they'll make that decision. 

**X6/11/29**

My child has been stolen.

**Author's Note:**

> Keebs getting stolen is canon you can't stop me  
> Anyway hope it was enjoyable, would love to hear any other backstory headcanons people might have. I figured 'be nice let the professor be a Good Guy' this time instead of 'haha the robot doesn't know his backstory is way more depressing with human context'.  
> maybe we're lucky and this is a vr au timeline and everything is okay after


End file.
